Primary producer communities are often growth-limited by essential nutrients such as nitrogen (N) and phosphorus (P). The magnitude of limitation and whether N, P, or both elements are limiting autotroph growth depends on the supply and ratios of these essential nutrients. Previous studies identified single, serial or co-limitation as predominant limitation outcomes in autotroph communities by factorial nutrient additions. Little is known about potential consequences of such scenarios for herbivores and whether their growth is primarily affected by changes in autotroph quantity or nutritional quality. We grew a community of phytoplankton species differing in various food quality aspects in experimental microcosms at varying N and P concentrations resulting in three different N:P ratios. At carrying capacity, N, P, both nutrients or none were added to reveal which nutrients were limiting. The nutrient supplied communities were fed to the generalist herbivorous rotifer Brachionus calyciflorus to investigate how changing phytoplankton biomass and community composition affect herbivore abundance. We found phytoplankton being growth-limited either by N alone (single limitation) or serially, i.e. primarily by N and secondarily by P, altering available food quantity for rotifers. Rotifer growth showed a different response pattern compared to phytoplankton, suggesting that apart from food quantity food quality aspects played a substantial role in the transfer from primary to secondary production. The combined addition of N and P to phytoplankton had generally a positive effect on herbivore growth, whereas adding non-limiting nutrients had a rather detrimental effect probably due to stoichiometrically imbalanced food in terms of nutrient excess. Our experiment shows that adding various nutrients to primary producer communities will not always lead to increased autotroph and herbivore growth, and that differences between autotroph and herbivore responses under co-limiting conditions can be partly well explained by concepts of ecological stoichiometry theory.

Details on how the data were collected and analysed are given in the methods section of the publication.

The experiment was set up in two time periods, a first pre-growing phase, in which phytoplankton communities grew on media with different N and P concentrations until they reached the stationary phase (i.e. were nutrient limited), and a second phase, in which growth of the communities was monitored after the addition of different amounts of N, P or both nutrients. Shortly after nutrient additions to phytoplankton communities, each microcosm was subsampled and distributed into wells previously stocked with asexually reproducing females of rotifer Brachionus calyciflorus, and their population growth was measured over time.

In the first phase, we composed a phytoplankton community consisting of 6 species of different nutritional quality aspects. We grew this community on 3 different COMBO media (N&P reduced, N reduced or P reduced) and monitored  the biomass development in the microcosms using absorbance as proxy. When the stationary phase was reached (all nutrients taken up) we subsampled the microcosms to assess community composition.

In the second phase, we addedd N, P, both or none to the microcosms and further measured biomass development. When the stationary phase was reached (all nutrients taken up) we subsampled the microcosms to assess again community composition. In addition, shortly after the nutrient addition, we started a parallel expeirment using phytoplankton from each microcosm as food for rotifer B. calyciflorus. We stocked rotifers in wells, fed them daily an aliquot from the micorcosms, and tracked rotifers abundance over several days. Then, rotifers were fixed and counted again.

This repository contains the data for "How nitrogen and phosphorus supply to nutrient-limited autotroph communities affects herbivore growth: testing stoichiometric and co-limitation theory across trophic levels" by Redoglio, Andrea; Radtke, Kassandra; Sperfeld, Erik, 2022, Oikos, doi:10.1111/oik.09052.

This repository contains the raw data and R scripts for "How nitrogen and phosphorus supply to nutrient-limited autotroph communities affects herbivore growth: testing stoichiometric and co-limitation theory across trophic levels" by Redoglio, Andrea; Radtke, Kassandra; Sperfeld, Erik, 2022, Oikos.

1. Data files: definitions of column headers, explanations of abbreviations and units used

A) "Data_AlgaeCounts.csv" - Datafile used for R Script "R-script_Fig.4&S2_Stats-Fig.4&S2.R"

Orientation: columns

Column name: Medium
Definition: Medium categories at the beginning of the experiment
Code: N&P low = nitrogen and phosphorus reduced; N low = nitrogen reduced; P low = phosphorus reduced
Missing: no
Type: text

Column name: Day
Definition: day of the experiment when phytoplankton samples were collected and fixed for biovolume determination
Code: 22 = experimental day 22; 49 = experimental day 49
Missing: no
Type: numeric
Unit: unitless

Column name: Algae
Definition: species of phytoplankton in the community
Code:  Acutodesmus = Acutodesmus obliquus; Monoraphidium = Monoraphidium minutum; Synechococcus = Synechococcus elongatus; Anabaena = Anabaena variabilis; Cryptomonas = Cryptomonas ovata; Nannochloropsis = Nannochloropsis limnetica
Missing: no
Type: text

Column name: Dilution
Definition: dilution coefficient of Lugol-fixed phytoplankton samples
Code: 10 = 1:10 dilution; 100 = 1:100 dilution
Missing: no
Type: numeric
Unit: unitless

Column name: Biovol_mm3
Definition: average single cell biovolume (measured in mm^3) of the respective species  
Missing: no
Type: numeric
Unit: cubic millimeters

Column(s) name: C1 - C2  - C3
Definition: number of phytoplankton cells in one quadrant of the Sedgewick-Rafter counting chamber in treatment CONTROL replicate 1, replicate 2 and replicate 3, respectively
Missing: no
Type: numeric
Unit: unitless

Column(s) name: N1 - N2  - N3
Definition: number of phytoplankton cells in one quadrant of the Sedgewick-Rafter counting chamber in treatment NITROGEN replicate 1, replicate 2 and replicate 3, respectively
Missing: no
Type: numeric
Unit: unitless

Column(s) name: P1 - P2  - P3
Definition: number of phytoplankton cells in one quadrant of the Sedgewick-Rafter counting chamber in treatment PHOSPHORUS replicate 1, replicate 2 and replicate 3, respectively
Missing: no
Type: numeric
Unit: unitless

Column(s) name: NP1 - NP2  - NP3
Definition: number of phytoplankton cells in one quadrant of the Sedgewick-Rafter counting chamber in treatment NITROGEN&PHOSPHORUS replicate 1, replicate 2 and replicate 3, respectively
Missing: no
Type: numeric
Unit: unitless

B) "Data_OD.csv" - Datafile used for R Script "R-script_Fig.2&3_Stats-Tab.2,Fig. 3.R"

Orientation: columns

Column name: Medium
Definition: Medium categories at the beginning of the experiment
Code: N&P low = nitrogen and phosphorus reduced; N low = nitrogen reduced; P low = phosphorus reduced
Missing: no
Type: text

Column name: Treatment
Definition: Treatment categories from the 2nd phase of the experiment
Code: Ctrl = no nitrogen and phosphorus added; N = nitrogen added; P = phosphorus added; N&P =  nitrogen and phosphorus added
Missing: no
Type: text

Column name: Replikat
Definition: replicates (flasks) of phytoplankton communities
Code: 1 = replicate 1, 2 = replicate 2 and 3 = replicate 3
Missing: no
Type: numeric
Unit: unitless

Columns name: 1 to 49
Definition: day of the experiment when the OD measurement was taken, at an interval of 3 days (every 3rd day)
Type: numeric
Unit: absorbance (unitless)

C) "Data_RotiferAbundance_OD_Biovolume.csv" - Datafile used for R Script "R-scrpt_Fig.6&S3_Stats-Fig.6&S3,Tab.3,Tab.S2,Tab.S3.R" and "R-script_Fig.S1.R"

Orientation: columns

Column names: Medium
Definition: Medium categories at the beginning of the experiment
Code: N&P low = nitrogen and phosphorus reduced; N low = nitrogen reduced; P low= phosphorus reduced
Missing: no
Type: text

Column names: Treatment
Definition: Treatment categories from the 2nd phase of the experiment
Code: Ctrl = no nitrogen and phosphorus added; N = nitrogen added; P = phosphorus added; N&P =  nitrogen and phosphorus added
Missing: no
Type: text

Column names: Replikat
Definition: replicate of phytoplankton community used to feed rotifers microcosms
Code: 1 = replicate 1, 2 = replicate 2 and 3 = replicate 3
Missing: no
Type: numeric
Unit: unitless

Column names: 1 to 35
Definition: day of the experiment when the number of rotifers was counted, at an interval of 1 day (everyday)
Type: numeric
Missing: no
Unit: rotifer abundance per well (unitless)

Column names: Well_ID
Definition: identity of a replicate of micorcosms (wells with rotifers)
Code: 1 = replicate 1, 2 = replicate 2, 3 = replicate 3 and 4 = replicate 4
Type: numeric
Missing: no
Unit: unitless

Column names: OD
Definition: optical density of the phytoplankton community given as food measured on day 49 (final day of the experiment)
Type: numeric
Missing: no
Unit: absorbance (unitless)

Column names: d22_Acu
Definition: biovolume of Acutodesmus obliquus on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d22_Mon
Definition: biovolume of Monoraphidium minutum on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d22_Syn
Definition: biovolume of Synechococcus elongatus on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d22_Ana
Definition: biovolume of Anabaena variabilis on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)
 
Column names: d22_Cry
Definition: biovolume of Cryptomonas ovata on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d22_Nan
Definition: biovolume of Nannochloropsis limnetica on day 22
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d49_Acu
Definition: biovolume of Acutodesmus obliquus on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d49_Mon
Definition: biovolume of Monoraphidium minutum on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d49_Syn
Definition: biovolume of Synechococcus elongatus on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d49_Ana
Definition: biovolume of Anabaena variabilis on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)
 
Column names: d49_Cry
Definition: biovolume of Cryptomonas ovata on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: d49_Nan
Definition: biovolume of Nannochloropsis limnetica on day 49
Type: numeric
Missing: no
Unit: mm3/ mL (cubic millimeters per milliliter)

Column names: OD_d35
Definition: optical density of the phytoplankton community given as food measured on day 35 (final day of the rotifers experiment)
Type: numeric
Missing: no
Unit: absorbance (unitless)

D) "Data_Excluded rotifers wells.csv" - Datafile used for R Script "R-script_Fig.S1.R"

Orientation: columns

Column name: medium
Definition: Medium categories at the beginning of the experiment
Code: N&P low = nitrogen and phosphorus reduced; N low = nitrogen reduced; P low= phosphorus reduced
Missing: no
Type: text

Column name: treatment
Definition: Treatment categories from the 2nd phase of the experiment
Code: Ctrl = no nitrogen and phosphorus added; N = nitrogen added; P = phosphorus added; N&P =  nitrogen and phosphorus added
Missing: no
Type: text

Column name: excluded
Definition: number of wells (microcosms with rotifers) excluded from the statistical analysis
Missing: no
Type: numeric
Unit: unitless

2. Statistics and data analyses

Statistical analysis and visualization of the data can be found in the R scripts.

All statistical analyses and tests of their assumptions were performed using the statistical software R, version 4.0.2. Data were visualized using the ggplot2 package. General linear models (lm function) were used to test the effects of medium and treatment on phytoplankton biomass on days 21 and 49, estimated by OD measurements. Tukey’s HSD post-hoc tests were used to assess differences between treatments within each media group. Post-hoc tests were applied for each medium separately to readily identify predicted factorial limitation scenarios in phytoplankton (Fig. 1). For the analyses of rotifer abundance, we excluded wells in which the rotifers died out or did not grow (exclusion criterion: cumulative daily counts were less than the number of counting days multiplied by the starting density, i.e. cumulative counts < 24 = 12 d × 2 rotifers, Fig. S1a) to avoid introducing a potential bias likely caused by stochastic extinction events (see above). Furthermore, there was no clear relationship between the number of excluded wells and media or treatment (Fig. S1b), suggesting no systematic effects of medium or treatment on mortality. Linear mixed-effects (LME) models were applied to test the fixed effects of phytoplankton growth medium and nutrient addition treatment on rotifer abundance counted at day 34 (alive) and at day 35 (Lugol-fixed) using the lmer function of the lme4 package (model 1). Phytoplankton replicate identity was used as random effect to account for potential non-treatment effects by using multiple rotifer wells per phytoplankton community replicate. P-values for ANOVA tables were computed using the anova function of the lmerTest package. Marginal and conditional R2 were calculated using the r2_nakagawa function of the performance package to assess the variance explained by fixed effects and the entire model, respectively. Multiple comparisons between treatments within each media category were conducted using the glht function of the multcomp package with p-values adjusted according to Holm, using log-transformed abundance data to mitigate heteroscedasticity. We applied additional LME models to investigate whether phytoplankton biomass and biovolume of particular phytoplankton species have a significant effect on rotifer abundance and how much variance these fixed variables explained. We added OD measured at the end of the rotifer experiment (day 35) as sole explanatory variable or together with the fixed factor medium to the models (model 2 and 3, respectively). We explored the role of biovolume of particular phytoplankton species with LME models. Because only green algae showed substantial treatment differences in the second phase after nutrient addition, we used the sum of biovolumes of the two green algae (A. obliquus and M. minutum) on day 49 as fixed variable in the model together with the fixed factor medium (model 4).